A Statistical and Analytical Method for Detecting Tiny Deformations in High Resolution 3D Laser Scanning Data

Terrestrial 3Dlaser scan data has high accuracy,high spatial resolution and can change thetraditional single-point observation mode.The traditional point measurement can be transformed intoshape measurement.Making it possible to detect tiny deformations in deformation monitoring datawhen the deformation is smaller than the measurement accuracy with mathematical and statisticalmethods.In this paper,we propose a statistical and analytical means to detect tiny deformations withhigh resolution 3Dlaser scanning data.We propose point cloud registration based on reflectance imageand BaySAC,and accurately registering it with ICP.Fitting can improve the overall analytical accura-cy,but the gross error is the main factor impacting fitting precision.We remove it with RANSAC al-gorithm.By eliminating random errors,revising registration errors,we ultimately arrive at the pre-cise deformation.The results from an experiment with building and subway tunnel data show that themethod is effective.